ME. J. G. BAKETJ OX LILIACE-E. 353 



to a large extent, run parallel with one another — Hyaemtheae, bulbs 

 of the gamophyllous subseries with racemose inflorescence, an- 

 swering to Scilleae, in which the segments of the perianth are 

 free, Hemerocallidese in the same way to Anthericese, Millea) to 

 Aliiese, and the other tribes in each of the two subseries less 

 closely. This is a general outline of what I believe to be the 

 most natural classification of the order; and in this paper all the 

 known genera and species of the gamophyllous capsular series 

 are reviewed, with the exception of the tribe Aloinese, which has 

 formed the subject of a beautifully illustrated monograph by 

 Prince Salm-Dyck. 



Of the genera, I believe I have had an opportunity of exa- 



mining a more or less perfect specimen of every one that has 

 been proposed. I have kept np all for which I thought I was 

 able to find any clearly definable structural individuality ; but 

 even as compared with the * Enumeratio,' I have felt bound to 

 erase a good many of the small ones from the list, finding, as I 

 proceeded, that it was quite impossible to do otherwise and at 

 the same time characterize genera with reasonable clearness. I 

 must not pass this point without acknowledging my obligations 

 to the fragment of Salisbury's * Genera Plantarum * which Dr. 

 J. E. Gray has lately so liberally printed and circulated. This 

 order seems to have been a particular favourite with that author; 

 and his researches upon it were made at a time when the cultiva- 

 tion of these plants was at its highest point of popularity. His 

 planning-out and definition of tlie genera show great care and 

 acuteuess ; but in circumscribing tliem he went upon a track the 

 direct opposite of that which I have followed. It will be seen 

 that, although I have only adopted one of his genera as a genus, 

 I have used many of his names and groups for subordinate divi- 

 sions. Indeed I may say that I have felt it only due to the 

 sterling merits of the work, and of a botanist who left behind 

 him no adequate memorial of his ability and industry, to incor- 

 porate as many of his groups and names as I could possibly 

 include. The total number of genera which I have defined is 

 26, and of species about 220, being an average of about eight 

 species to a genus. In this portion of the capsular series, a 

 condition of things quite different from what we see in other 

 parts of the order, both large genera and good monotypic genera 

 are comparatively few in number, the rule being genera of small 

 or moderate size, in most of the tribes clearly bounded, with 



